Skip to content

[ICML 2022] Active learning of AI models on ECG data with SoQal

Notifications You must be signed in to change notification settings

danikiyasseh/SoQal

Repository files navigation

Active Learning of Cardiac Signals with SoQal

SoQal is framework that allows a network to dynamically decide, upon acquiring an unlabelled data point, whether to request a label for that data point from an oracle or to pseudo-label it instead. It can reduce a network's dependence on an oracle (e.g., physician) while maintaining its strong predictive performance.

This repository contains a PyTorch implementation of SoQal. For details, see SoQal: Selective Oracle Questioning for Consistency Based Active Learning of Cardiac Signals. [ICML paper] [blogpost] [video]

Requirements

The SoQal code requires

  • Python 3.6 or higher
  • PyTorch 1.0 or higher

Datasets

Download

The datasets can be downloaded from the following links:

  1. PhysioNet 2015
  2. PhysioNet 2017
  3. Cardiology

Pre-processing

In order to pre-process the datasets appropriately for SoQal, please refer to the following repository

Training

To train the model(s) in the paper, run this command:

python run_experiments.py

Evaluation

To evaluate the model(s) in the paper, run this command:

python run_experiments.py

About

[ICML 2022] Active learning of AI models on ECG data with SoQal

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages